Although recent studies have shown the importance of complex genomic rearrangements in prostate cancer initiation and progression, the mechanisms that generate these rearrangements are unknown. Our long-term goal is to understand the causes and consequences of these complex rearrangements in prostate and other cancers. The objective here is to determine the mechanism by which multiple, coordinate, double strand breaks (DSB) likely involved in generating these complex rearrangements can form. Our central hypothesis is that initiation of androgen-induced transcriptional programs in neoplastic prostate cells involves the formation of highly complex topological loops that are mediated in part by topoisomerase II beta (TOP2B); dysfunction of this enzyme can produce multiple simultaneous breaks that can recombine to form complex genomic rear- rangements. We have formulated this hypothesis on the basis of our strong preliminary data showing andro- gen-induced, TOP2B-mediated DSB that are recombinogenic and arise near rearrangement breakpoints ob- served in human prostate cancers. The rationale for this work is that understanding these mechanisms can inform strategies to prevent formation of these rearrangements, potentially allowing cancer prevention and interception, and may also spur development of rearrangement-based biomarkers for prostate cancer risk stratification. We will test our hypothesis through three specific aims: 1) Determine the role of TOP2B in chromosomal conformational remodeling during initiation of androgen receptor (AR) mediated transcriptional programs. 2) Identify the role of TOP2B binding and catalytic activity in generating androgen-induced double strand breaks and recombination/rearrangement events. 3) Determine the relevance and utility of a TOP2B-related rearrangement signature in human prostate cancer. To facilitate these aims, we have innovated and/or established feasibility in our hands for: i) robust methods for genetic and pharmacologic inhibition of TOP2B; ii) well defined prostate cancer cell line model systems to study androgen receptor signaling; iii) identification o rearrangement breakpoints in human prostate cancers with low-pass paired-end sequencing; iv) methods for measuring TOP2B-mediated chromosomal looping interactions, gene expression changes, DSB, and recombination events; and iv) powerful statistical methods for assessing the spatial correlations of these TOP2B- related events with each other and with rearrangement breakpoints identified in human cancers. Using such correlations with data collected here and with publically available data, we will be able to assess whether a TOP2B/AR rearrangement signature has relevance to human prostate cancer rearrangements and whether it can have utility for prostate cancer risk stratification. The approach is innovative and highly significant because it uses powerful new experimental and computational methodologies to test an innovative hypothesis explaining the formation of highly complex, context-specific rearrangements observed in human prostate cancer. This understanding can ultimately inform approaches for prostate cancer prevention and interception.